Crash prone 'modern roundabouts'

[kalvado]

Why do you have such a grudge against roundabouts? Did your hometown get an undeserving roundabout like mine did?

I don't have any particular grudge towards multi-lane roundabouts, but you can't honestly deny that multi-lane roundabouts do have a lot more crashes than single-lane roundabouts.

Can't disagree, and it's not wrong since it's backed up with stats, but this is a 117-page thread about the same thing. I think I get it after 5 pages.

That means you're smarter than most DOT engineers!

[jakeroot]

Why do you have such a grudge against roundabouts? Did your hometown get an undeserving roundabout like mine did?

I don't have any particular grudge towards multi-lane roundabouts, but you can't honestly deny that multi-lane roundabouts do have a lot more crashes than single-lane roundabouts.

Can't disagree, and it's not wrong since it's backed up with stats, but this is a 117-page thread about the same thing. I think I get it after 5 pages.

That means you're smarter than most DOT engineers!

Bingo.

If it's so painfully obvious to us "unordained", then why do we keep installing them and keep telling everyone how they reduce crashes? When they do...if they're simple 1x1 or 1x2 roundabouts. These giant 2x2 or 2x3 roundabouts do not reduce crashes, and never have.

In fairness, I'm a huge proponent of double permissive left turns, I started a huge thread on them years ago. Even I will concede they generally increase crashes, though only by a small amount. But that's the key: "small amount", huge multi-lane roundabouts often see double or triple the number of crashes as before. As Daniel Tosh said, it's not a stereotype if it's always true...then it becomes law. And the law is: 2x2 or 2x3 roundabouts have tons of crashes, period. Even in Maryland.

[Rothman]

Meh.  Although minor crash rates increased at 2-lane roundabouts initially that were installed in the Albany area of NY, such have subsided quite dramatically and rather quickly and are much preferrable to the previous traffic lights.

Any roundabout that continues to be an issue should surely be reassessed, but the blanket statement that all 2-lane roundabouts are dangerous is simply unfounded.

I've found tradephoric's snapshots in time and focuses on scattershot problem roundabouts as support for the idea that all roundabouts are problematic to be specious reasoning.

[kalvado]

Meh.  Although minor crash rates increased at 2-lane roundabouts initially that were installed in the Albany area of NY, such have subsided quite dramatically and rather quickly and are much preferrable to the previous traffic lights.

Any roundabout that continues to be an issue should surely be reassessed, but the blanket statement that all 2-lane roundabouts are dangerous is simply unfounded.

I've found tradephoric's snapshots in time and focuses on scattershot problem roundabouts as support for the idea that all roundabouts are problematic to be specious reasoning.

Malta just loves all those trailer rollovers! And those are not even crashes, right?

[tradephoric]

Meh.  Although minor crash rates increased at 2-lane roundabouts initially that were installed in the Albany area of NY, such have subsided quite dramatically and rather quickly and are much preferrable to the previous traffic lights.

Nice story, now show me the data. 

[jakeroot]

Any roundabout that continues to be an issue should surely be reassessed, but the blanket statement that all 2-lane roundabouts are dangerous is simply unfounded.

That's not fair, no one has claimed they are dangerous. In fact, that's the whole issue: 2x2 or 2x3 (or 3x3) multi-lane roundabouts are not very dangerous. They usually have very low injury rates, which is why they continue to be installed. However, property-damage-only collisions skyrocket almost universally.

We have talked about the dangers of central island structures, but that's another topic within this thread.

[Rothman]

Any roundabout that continues to be an issue should surely be reassessed, but the blanket statement that all 2-lane roundabouts are dangerous is simply unfounded.

That's not fair, no one has claimed they are dangerous. In fact, that's the whole issue: 2x2 or 2x3 (or 3x3) multi-lane roundabouts are not very dangerous. They usually have very low injury rates, which is why they continue to be installed. However, property-damage-only collisions skyrocket almost universally.

We have talked about the dangers of central island structures, but that's another topic within this thread.

The property-only damage incidents do reduce over time on properly installed roundabouts.

[kalvado]

Any roundabout that continues to be an issue should surely be reassessed, but the blanket statement that all 2-lane roundabouts are dangerous is simply unfounded.

That's not fair, no one has claimed they are dangerous. In fact, that's the whole issue: 2x2 or 2x3 (or 3x3) multi-lane roundabouts are not very dangerous. They usually have very low injury rates, which is why they continue to be installed. However, property-damage-only collisions skyrocket almost universally.

We have talked about the dangers of central island structures, but that's another topic within this thread.

My bigger problem is that there is no understanding behind designs.
Just remember this picture:

It was used a lot to "prove" that collisions should go down universally and unconditionally. Reality is pretty predictable, though.

[jakeroot]

Any roundabout that continues to be an issue should surely be reassessed, but the blanket statement that all 2-lane roundabouts are dangerous is simply unfounded.

That's not fair, no one has claimed they are dangerous. In fact, that's the whole issue: 2x2 or 2x3 (or 3x3) multi-lane roundabouts are not very dangerous. They usually have very low injury rates, which is why they continue to be installed. However, property-damage-only collisions skyrocket almost universally.

We have talked about the dangers of central island structures, but that's another topic within this thread.

The property-only damage incidents do reduce over time on properly installed roundabouts.

Generally thinking, I'd think so, but this thread very much suggests otherwise. The intersections in Malta are no exception, if this study from the FHWA is to be believed. Failure to yield appears to be a big issue at the Dunning St / US-9 / NY-67 roundabout.

[Rothman]

Any roundabout that continues to be an issue should surely be reassessed, but the blanket statement that all 2-lane roundabouts are dangerous is simply unfounded.

That's not fair, no one has claimed they are dangerous. In fact, that's the whole issue: 2x2 or 2x3 (or 3x3) multi-lane roundabouts are not very dangerous. They usually have very low injury rates, which is why they continue to be installed. However, property-damage-only collisions skyrocket almost universally.

We have talked about the dangers of central island structures, but that's another topic within this thread.

The property-only damage incidents do reduce over time on properly installed roundabouts.

Generally thinking, I'd think so, but this thread very much suggests otherwise. The intersections in Malta are no exception, if this study from the FHWA is to be believed. Failure to yield appears to be a big issue at the Dunning St / US-9 / NY-67 roundabout.

Any roundabout that continues to be an issue should surely be reassessed, but the blanket statement that all 2-lane roundabouts are dangerous is simply unfounded.

That's not fair, no one has claimed they are dangerous. In fact, that's the whole issue: 2x2 or 2x3 (or 3x3) multi-lane roundabouts are not very dangerous. They usually have very low injury rates, which is why they continue to be installed. However, property-damage-only collisions skyrocket almost universally.

We have talked about the dangers of central island structures, but that's another topic within this thread.

The property-only damage incidents do reduce over time on properly installed roundabouts.

Generally thinking, I'd think so, but this thread very much suggests otherwise. The intersections in Malta are no exception, if this study from the FHWA is to be believed. Failure to yield appears to be a big issue at the Dunning St / US-9 / NY-67 roundabout.

Thia thread, as it has been pointed out again and again, suffers from an overwhelming selection bias.  As the foreword of that study points out, the yielding problem only exists for "some" multilane roundabouts.

For every Malta, there are more successful multilane roundabouts, such as those that I've pointed out before on NY 85 in Slingerlands.  Even the NY 85/NY 140 roundabout, which had frequent crashes when it opened, has had incidents greatly reduced over time as expected by the current true state-of-the-art on roundabouts.

[jeffandnicole]

Just remember this picture:

It was used a lot to "prove" that collisions should go down universally and unconditionally. Reality is pretty predictable, though.

This picture proves that anti-car and/or pro-roundabout people like to exaggerate hazards.

For pedestrian hazards: The first picture shows that there are 3 points of conflict, I'm guessing due to left, straight and right movements. While the graph isn't detailed, let's assume there's 1 lane per movement. For the roundabout, it only shows 1 potential collision point, despite there clearly being 2 lanes.

For vehicle hazards, it shows every possible collision point from two motorists entering the intersection at a time, and of course one would need to do so illegally. For the roundabout, there's an assumption made that no vehicle will enter or change lanes illegally.

The first pic also shows that rear-end crashes approaching an intersection are conflict points. Yet the second intersection doesn't show rear-end crashes being a possibility.  Actually, there's 2 red dots at what would be the stop line. Maybe it's claiming traffic just beginning a turn will hit each other. Yet traffic entering the roundabout apparently are perfect drivers and will never sideswipe each other.

[kalvado]

Just remember this picture:

It was used a lot to "prove" that collisions should go down universally and unconditionally. Reality is pretty predictable, though.

This picture proves that anti-car and/or pro-roundabout people like to exaggerate hazards.

For pedestrian hazards: The first picture shows that there are 3 points of conflict, I'm guessing due to left, straight and right movements. While the graph isn't detailed, let's assume there's 1 lane per movement. For the roundabout, it only shows 1 potential collision point, despite there clearly being 2 lanes.

For vehicle hazards, it shows every possible collision point from two motorists entering the intersection at a time, and of course one would need to do so illegally. For the roundabout, there's an assumption made that no vehicle will enter or change lanes illegally.

The first pic also shows that rear-end crashes approaching an intersection are conflict points. Yet the second intersection doesn't show rear-end crashes being a possibility.  Actually, there's 2 red dots at what would be the stop line. Maybe it's claiming traffic just beginning a turn will hit each other. Yet traffic entering the roundabout apparently are perfect drivers and will never sideswipe each other.

Great analysis... Too bad it's totally irrelevant.
In both cases pedestrian has to deal with the same number of vehicles coming in within the same timeframe.
In both cases driver has to deal with same car-to-car conflicts... except for there are more conflicts for roundabouts.

As for number of conflict points... what's more risky, a single intersection of 2 arterials with 40k traffic in each, or 4x4 grid where each intersection is 10 + 10 k?

[tradephoric]

Thia thread, as it has been pointed out again and again, suffers from an overwhelming selection bias.  As the foreword of that study points out, the yielding problem only exists for "some" multilane roundabouts.

For every Malta, there are more successful multilane roundabouts, such as those that I've pointed out before on NY 85 in Slingerlands.  Even the NY 85/NY 140 roundabout, which had frequent crashes when it opened, has had incidents greatly reduced over time as expected by the current true state-of-the-art on roundabouts.

The problem roundabouts with high number of crashes are the complex 2x2 and 2x3 roundabouts.  None of the NY roundabouts you cited are complex roundabouts that have at least 2 entering and exiting thru lanes at each approach.  You may be biased from improper selection of a cohort that does not closely represent the greater population for which the study aims to be applicable... AKA selection bias. 

I'd love it if you cited a complex roundabout in NY with a low crash rate.  Then we could start to look at how it is designed and see what separates it out from all the other complex roundabouts with high crash rates.  I'll be waiting.

[kalvado]

Thia thread, as it has been pointed out again and again, suffers from an overwhelming selection bias.  As the foreword of that study points out, the yielding problem only exists for "some" multilane roundabouts.

For every Malta, there are more successful multilane roundabouts, such as those that I've pointed out before on NY 85 in Slingerlands.  Even the NY 85/NY 140 roundabout, which had frequent crashes when it opened, has had incidents greatly reduced over time as expected by the current true state-of-the-art on roundabouts.

The problem roundabouts with high number of crashes are the complex 2x2 and 2x3 roundabouts.  None of the NY roundabouts you cited are complex roundabouts that have at least 2 entering and exiting thru lanes at each approach.  You may be biased from improper selection of a cohort that does not closely represent the greater population for which the study aims to be applicable... AKA selection bias. 

I'd love it if you cited a complex roundabout in NY with a low crash rate.  Then we could start to look at how it is designed and see what separates it out from all the other complex roundabouts with high crash rates.  I'll be waiting.

Malta my have a different problem though, with an overabundance of events like these:
https://cbs6albany.com/news/local/tractor-trailer-rollover-stalls-traffic-in-malta-dunning-street-saratoga-county-
Given very highway exit-centric area layout, closing that road for most of the day is a pretty significant event

[tradephoric]

How about 2 rollovers exactly one week apart at a roundabout in Marengo, Illinois.  Story published 10/1/24.

Semi-truck rolls over in roundabout in Marengo week after semi-truck rollover in same location
https://www.lakemchenryscanner.com/2024/10/01/semi-truck-rolls-over-in-roundabout-in-marengo-week-after-semi-truck-rollover-in-same-location/

[tradephoric]

Still waiting for someone to cite a complex roundabout with a low crash rate.  Where's your data Rothman?

[tradephoric]

A bunch of crashes at a newly built roundabout at Irvington & Patton in Houston.  There is plenty of signage warning drivers they are approaching a roundabout, but it seems like drivers just don't see the roundabout coming especially at night.  It appears to be a concrete central island and it just seems to blend in with the roadway. 

https://abc13.com/post/northside-roundabout-irvington-patton-stirs-controversy-after-houston-mayor-john-whitmire-decides-future/15488879/

[tradephoric]

The roundabout at Wickham Road and Lake Andrew Drive has become the most crash prone intersection in Brevard County, Florida (population 643k).  It averages 59 crashes per year with the next closest intersection in the county (Babcock Street and Palm Bay Road in Palm Bay) averaging 54.4 crashes. 

Viera roundabout crashes on the rise. But data shows it's safer than you might think
https://www.floridatoday.com/story/news/2024/05/21/viera-roundabout-traffic-crashes-florida-space-coast/73345213007/

[tradephoric]

Roundabouts curb severe crashes, study shows, even if fender-benders increase
https://www.toledoblade.com/local/transportation/2024/06/29/roundabouts-curb-severe-crashes-fender-benders-increase/stories/20240627115

A study by the Toledo Metropolitan Area Council of Governments analyzed 55 roundabouts in the region.  They found that there were 588 crashes at the roundabout locations during the three years before construction, and 769 crashes the three years after.  Total crashes increased around 30%.

One bad performing roundabout in a region can completely skew the results though... and we should try to stop building these "bad" roundabouts.  The roundabout at Detroit/Berdan/Cherry opened in 2015 and crashes increased from just 30 between 2011 and 2013 to 258 between 2016 and 2018.  That's a 760% increase in crashes.  The article doesn't even cite the number of injury crashes that occurred, but it did estimate the annual cost of damage and injuries from crashes there rose from $913,331 to $3,257,521 (a 256% increase).   

In the case of these 55 Ohio roundabouts studied, there were 4 reported fatalities before the roundabouts were built and 2 fatalities after the roundabouts were built.  So the worst type of crashes (a fatal crash) did drop in half.  That's a good thing.

[tradephoric]

Total crashes at these Ohio roundabouts would have actually dropped had the region not build that Detroit/Berdan/Cherry roundabout.  Before crashes would have dropped to 558 (from 588) and after crashes would have dropped to 511 (from 769).  Total crashes at the roundabout in the region would have dropped by 8% instead of rose by 30%.  Hopefully lesson learned... stop building these complex 2x2 and 2x3 roundabouts. 

[JCinSummerfield]

I used to traverse the Detroit/Berdan/Cherry roundabout when commuting to my old job.  You kept your head on a swivel and prayed to get through the roundabout without someone turning from the wrong lane into your car.

[tradephoric]

Bluemont, Fourth St. roundabout most dangerous intersection in Manhattan
https://themercury.com/news/bluemont-fourth-st-roundabout-most-dangerous-intersection-in-manhattan/article_66583ed0-a8f2-11ef-a38b-c3e9b5605479.html

[tradephoric]

The City of Fishers built a new complex 2x2 roundabout to replace the Michigan Left intersection at 96th Street and Andersonville.  Drivers are complaining about how long it's taking to make it through the intersection now that the roundabout is complete.

Carmel built a bunch of single lane and simple (1x2) multi-lane roundabouts over the past 25 years.  But the need for complex 2x2 and 2x3 roundabouts was largely avoided since the City decided to grade separate their major corridors (Meridian Street & Keystone Pkwy).  Now Fishers is getting in on the fun, but unlike Carmel they are not grade-separating their major corridor along Allisonville Road.  They just completed a complex 2x2 roundabout at 96th and Allisonville and they are planning an even more complex/busy roundabout at 116th and Allisonville Road this Spring.  These are the type of roundabouts where you see triple digit crashes per year.  Not all roundabouts are created equal, and Fishers is really pushing the limits to how big and complex you want to build these. 

[cjw2001]

As the video mentioned, part of the problem is increased traffic from the current major reconstruction of the nearby I 465 and I 69 interchange which has some drivers detouring onto local roads.  That said, some of the issue is also the Indianapolis driver base which is much less familiar with roundabouts than their Hamilton county counterparts.  I would expect this to improve with time as those drivers become used to the intersection.  I've driven through the new roundabout many times and have had zero issues with back ups.  But "the intersection is working well most of the time" doesn't make for good clicks for the local news media.

Also the new roundabout currently under construction at 146th and Allisonville is a Hamilton County project and includes grade separation plus the expansion of a nearby bridge over the White River.

https://hamiltoncounty.in.gov/1238/146th-St-Allisonville-Rd-Interchange

https://www.flickr.com/photos/hamiltoncountyhighway/54124369680/in/album-72157686645026180

[mrsman]

Can someone explain the terminology used for roundabouts like 1X2 and 2X2 and 2X3?  What does each number represent, the number of lanes entering the roundabout or the number of lanes in the roundabout itself?